Introduction
Modelling, a critical component within the purview of Social Cognitive Theory, doth denote alterations, be they behavioural, cognitive, or affective, emanating from the observation of one or more exemplars (Rosenthal & Bandura, 1978; Schunk, 1987, 1998; Zimmerman, 1977). In times past, modelling was oft conflated with mere imitation, yet modelling doth encompass a more comprehensive concept. A historical exposition shall forthwith be presented, thereby providing a backdrop against which the import of modelling research, as conducted by Bandura and divers others, may be duly appreciated.
Theories of Imitation
Throughout history, scholars have regarded imitation as a significant means of transmitting behaviours (Rosenthal & Zimmerman, 1978). The ancient Greeks employed the term 'mimesis' to denote learning through observation of the actions of others, and of abstract models exemplifying literary and moral styles. Other perspectives on imitation relate it to instinct, development, conditioning, and instrumental behaviour.
| View | Assumptions |
|---|---|
| Instinct | Observed actions elicit an instinctive drive to copy those actions. |
| Development | Children imitate actions that fit with existing cognitive structures. |
| Conditioning | Behaviours are imitated and reinforced through shaping. Imitation becomes a generalised response class. |
| Instrumental behaviour | Imitation becomes a secondary drive through repeated reinforcement of responses matching those of models. Imitation results in drive reduction. |
Instinct
At the commencement of the twentieth century, the prevailing scientific view was that individuals possessed a natural instinct to imitate the actions of others (James, 1890; Tarde, 1903). James posited that imitation was largely responsible for socialisation, but he did not elucidate the process by which imitation occurs. McDougall (1926) restricted his definition of imitation to the instinctive copying by one person of the actions of another.
Behaviourists rejected the notion of instinct (and thus it became discarded) because it presumed the existence of an internal drive, and possibly a mental image, intervening between a stimulus (action of another person) and response (copying of that action). Watson (1924) believed that people’s behaviours labelled 'instinctive' resulted largely from training and therefore were learned.
Development
Piaget (1962) offered a different perspective on imitation. He believed that human development involved the acquisition of schemes (schemas), or cognitive structures that underlie and enable organised thought and action (Flavell, 1985). Thoughts and actions are not synonymous with schemes; they are overt manifestations of schemes. Schemes available to individuals determine how they react to events. Schemes reflect prior experiences and comprise one’s knowledge at any given time.
Schemes presumably develop through maturation and experiences slightly more advanced than one’s existing cognitive structures. Imitation is restricted to activities corresponding to existing schemes. Children may imitate actions they understand, but they should not imitate actions incongruent with their cognitive structures. Development, therefore, must precede imitation.
This view severely limits the potential of imitation to create and modify cognitive structures. Further, there is little empirical support for this developmental position (Rosenthal & Zimmerman, 1978). In an early study, Valentine (1930b) found that infants could imitate actions within their capabilities that they had not previously performed. Infants showed a strong tendency to imitate unusual actions commanding attention. The imitation was not always immediate, and actions often had to be repeated before infants would imitate them. The individual performing the original actions was important: Infants were most likely to imitate their mothers. These and results from subsequent research demonstrate that imitation is not a simple reflection of developmental level but rather may serve an important role in promoting development (Rosenthal & Zimmerman, 1978).
Conditioning
Conditioning theorists construe imitation in terms of associations. According to Humphrey (1921), imitation is a type of circular reaction in which each response serves as a stimulus for the next response. A baby may commence crying (response) because of a pain (stimulus). The baby hears its own crying (auditory stimulus), which then becomes a stimulus for subsequent crying. Through conditioning, small reflex units form progressively more complex response chains.
Skinner’s (1953) operant conditioning theory treats imitation as a generalised response class (Chapter 3). In the three-term contingency (S D → R → S R), a modelled act serves as the S D (discriminative stimulus). Imitation occurs when an observer performs the same response (R) and receives reinforcement (S R). This contingency becomes established early in life. For example, a parent makes a sound ('Dada'), the child imitates, and the parent delivers reinforcement (smile, hug). Once an imitative response class is established, it can be maintained on an intermittent reinforcement schedule. Children imitate the behaviours of models (parents, friends) as long as the models remain discriminative stimuli for reinforcement.
A limitation of this view is that one can imitate only those responses one can perform. In fact, much research demonstrates that diverse types of behaviours can be acquired through observation (Rosenthal & Zimmerman, 1978). Another limitation concerns the need for reinforcement to produce and sustain imitation. Research by Bandura and others shows that observers learn from models in the absence of reinforcement to models or observers (Bandura, 1986). Reinforcement primarily affects learners’ performance of previously learned responses rather than new learning.
Instrumental Behavior
Miller and Dollard (1941) proposed an elaborate theory of imitation, or matched-dependent behaviour, which contends that imitation is instrumental learned behaviour because it leads to reinforcement. Matched-dependent behaviour is matched to (the same as) that of the model and depends on, or is elicited by, the model’s action.
Miller and Dollard believed that initially the imitator responds to behavioural cues in trial-and-error fashion, but eventually the imitator performs the correct response and is reinforced. Responses performed by imitators previously were learned.
This conception of imitation as learned instrumental behaviour was an important advance, but it has problems. Like other historical views, this theory postulates that new responses are not created through imitation; rather, imitation represents performance of learned behaviours. This position cannot account for learning through imitation, for delayed imitation (i.e., when imitators perform the matching responses some time after the actions are performed by the model), or for imitated behaviours that are not reinforced (Bandura & Walters, 1963). This narrow conception of imitation restricts its usefulness to imitative responses corresponding closely to those portrayed by models.
Functions of Modelling
Bandura (1986) did distinguish three key functions of modelling: response facilitation, inhibition/disinhibition, and observational learning.
| Function | Underlying Process |
|---|---|
| Response facilitation | Social prompts do create motivational inducements for observers to model the actions (“going along with the crowd”). |
| Inhibition and disinhibition | Modelled behaviours do create expectations in observers that they shall experience similar consequences should they perform the actions. |
| Observational learning | Processes include attention, retention, production, and motivation. |
Response Facilitation
People do learn many skills and behaviours that they do not perform because they lack motivation to do so. Response facilitation doth refer to modelled actions that serve as social prompts for observers to behave accordingly. Consider an elementary teacher who hath set up an attractive display in a corner of the classroom. When the first students enter in the morning, they spot the display and immediately go to look at it. When other students enter the room, they see a group in the corner, so they, too, move to the corner to see what everyone is looking at. Several students together serve as a social prompt for others to join them, even though the latter may not know why the others are gathered.
Response facilitation effects are common. Hast thou ever seen a group of people looking in one direction? This can serve as a cue for thee to look in the same direction. Newcomers at meetings of volunteer groups may watch with interest as a basket is passed for donations. If most people put in a dollar, that serves as a signal that a dollar is an acceptable donation. Note that response facilitation doth not reflect true learning because people already know how to perform the behaviours. Rather, the models serve as cues for observer’s actions. Observers gain information about the appropriateness of behaviour and may be motivated to perform the actions if models receive positive consequences.
Response facilitation modelling may occur without conscious awareness. Chartrand and Bargh (1999) did find evidence for a Chameleon effect, whereby people nonconsciously mimic behaviours and mannerisms of people in their social environments. Simply perceiving behaviour may trigger a response to act accordingly.
Inhibition / Disinhibition
Observing a model can strengthen or weaken inhibitions to perform behaviours previously learned. Inhibition doth occur when models are punished for performing certain actions, which in turn doth stop or prevent observers from acting accordingly. Disinhibition doth occur when models perform threatening or prohibited activities without experiencing negative consequences, which may lead observers to perform the same behaviours. Inhibitory and disinhibitory effects on behaviour occur because the modelled displays convey to observers that similar consequences are probable if they perform the modelled behaviours. Such information also may affect emotions (e.g., increase or decrease anxiety) and motivation.
Teachers’ actions can inhibit or disinhibit classroom misbehaviour. Unpunished student misbehaviour may prove disinhibiting: Students who observe modelled misbehaviour not punished might start misbehaving themselves. Conversely, misbehaviour in other students may be inhibited when a teacher disciplines one student for misbehaving. Observers are more likely to believe that they, too, shall be disciplined if they continue to misbehave and are spotted by the teacher.
Inhibition and disinhibition are similar to response facilitation in that behaviours reflect actions people already have learned. One difference is that response facilitation generally involves behaviours that are socially acceptable, whereas inhibited and disinhibited actions often have moral or legal overtones (i.e., involve breaking rules or laws) and have accompanying emotions (e.g., fears). Looting may occur during a riot or natural disaster if looters go unpunished, which disinhibits looting (an illegal act) in some observers.
Observational Learning
Observational learning through modelling doth occur when observers display new patterns of behaviour that, prior to exposure to the modelled behaviours, have a zero probability of occurrence even when motivation is high (Bandura, 1969). A key mechanism is the information conveyed by models to observers of ways to produce new behaviours (Rosenthal & Zimmerman, 1978). In the opening scenario, Donnetta needed to learn (or relearn) the correct procedure for hitting a backhand. Observational learning doth comprise four processes: attention, retention, production, and motivation (Bandura, 1986)
The first process is observer attention to relevant events so that they are meaningfully perceived. At any given moment one can attend to many activities. Characteristics of the model and the observer influence one’s attention to models. Task features also command attention, especially unusual size, shape, colour, or sound. Teachers often make modelling more distinctive with bright colours and oversized features. Attention also is influenced by perceived functional value of modelled activities. Modelled activities that observers believe are important and likely to lead to rewarding outcomes command greater attention. Students believe that most teacher activities are highly functional because they are intended to enhance student learning. Learners also are apt to believe that their teachers are highly competent, which enhances attention. Factors that promote the perception of model competence are modelled actions that lead to success and symbolic indicators of competence, such as one’s title or position.
| Process | Activities |
|---|---|
| Attention | Student attention is directed by physically accentuating relevant task features, subdividing complex activities into parts, using competent models, and demonstrating usefulness of modelled behaviours. |
| Retention | Retention is increased by rehearsing information to be learned, coding in visual and symbolic form, and relating new material to information previously stored in memory. |
| Production | Behaviours produced are compared to one’s conceptual (mental) representation. Feedback helps to correct deficiencies. |
| Motivation | Consequences of modelled behaviours inform observers of functional value and appropriateness. Consequences motivate by creating outcome expectations and raising self-efficacy. |
The second process is retention, which doth require cognitively organising, rehearsing, coding, and transforming modelled information for storage in memory. Observational learning doth postulate two modes of storing knowledge. A modelled display can be stored as an image, in verbal form, or both (Bandura, 1977b). Imaginal coding is especially important for activities not easily described in words; for example, motor skills performed so rapidly that individual movements merge into a larger organised sequence or act (e.g., golf swing). Much cognitive skill learning doth rely upon verbal coding of rules or procedures.
Rehearsal, or the mental review of information, doth serve a key role in the retention of knowledge (Chapter 5). Bandura and Jeffery (1973) did find benefits of coding and rehearsal. Adults were presented with complex-modelled movement configurations. Some participants coded these movements at the time of presentation by assigning to them numerical or verbal designators. Other participants were not given coding instructions but were told to subdivide the movements to remember them. In addition, participants either were or were not allowed to rehearse the codes or movements following presentation. Both coding and rehearsal enhanced retention of modelled events; individuals who coded and rehearsed showed the best recall. Rehearsal without coding and coding without rehearsal were less effective.
The third observational learning process is production, which involves translating visual and symbolic conceptions of modelled events into overt behaviours. Many simple actions may be learned by simply observing them; subsequent production by observers doth indicate learning. Rarely, however, are complex behaviours learned solely through observation. Learners often shall acquire a rough approximation of a complex skill by observing modelled demonstrations (Bandura, 1977b). They then refine their skills with practice, corrective feedback, and reteaching.
Problems in producing modelled behaviours arise not only because information is inadequately coded but also because learners experience difficulty translating coded information in memory into overt action. For example, a child may have a basic understanding of how to tie shoelaces but not be able to translate that knowledge into behaviour. Teachers who suspect that students are having trouble demonstrating what they have learned may need to test students in different ways.
Motivation, the fourth process, doth influence observational learning because people are more likely to engage in the preceding three processes (attention, retention, production) for modelled actions that they feel are important. Individuals form expectations about anticipated outcomes of actions based on consequences experienced by them and models (Bandura, 1997). They perform those actions they believe shall result in rewarding outcomes and avoid acting in ways they believe shall be responded to negatively (Schunk, 1987). Persons also act based on their values, performing activities they value and avoiding those they find unsatisfying, regardless of the consequences to themselves or others. People forgo money, prestige, and power when they believe activities they must engage in to receive these rewards are unethical (e.g., questionable business practices).
Motivation is a critical process of observational learning that teachers promote in various ways, including making learning interesting, relating material to student interests, having students set goals and monitor goal progress, providing feedback indicating increasing competence, and stressing the value of learning.
Cognitive Skill Learning
Observational learning doth expand the compass and velocity of learning beyond what might transpire through shaping (vide Chapter 3), where each response must be performed and reinforced. Modelled portrayals of cognitive skills are standard features in classrooms. In a common instructional sequence, a teacher explains and demonstrates the skills to be acquired, after which the students receive guided practice whilst the teacher doth check for student understanding. The skills are retaught if students experience difficulty.
When the teacher is satisfied that students possess a basic understanding, they may engage in independent practice whilst the teacher periodically monitors their work.
Many features of instruction incorporate models, and there is much research evidence showing that students of various ages learn skills and strategies by observing models (Horner, 2004; Schunk, 2008). Two especially germane applications of modelling to instruction are cognitive modelling and self-instruction.
Cognitive Modelling
Cognitive modelling doth incorporate modelled explanation and demonstration with verbalisation of the model’s thoughts and reasons for performing given actions (Meichenbaum, 1977). Coach Martin used cognitive modelling with Donnetta. In teaching division skills, a teacher might verbalise the following in response to the problem 27 ÷ 4:
First, I have to decide what number to divide 4 into. I take 27, start on the left, and move toward the right until I have a number the same as or larger than 4. Is 2 larger than 4? Nay. Is 27 larger than 4? Aye. So my first division will be 4 into 27. Now I need to multiply 4 by a number that will give an answer the same as or slightly smaller than 27. How about 5? . Nay, too small. Let’s try . Maybe. Let’s try . Nay, too large. So 6 is correct.
Cognitive modelling can include other types of statements. Errors may be built into the modelled demonstration to show students how to recognise and cope with them. Self-reinforcing statements, such as “I’m doing well,” also are useful, especially with students who encounter difficulties learning and doubt their capabilities to perform well.
Researchers have substantiated the useful role of cognitive modelling and shown that modelling combined with explanation is more effective in teaching skills than explanation alone (Rosenthal & Zimmerman, 1978). Schunk (1981) compared the effects of cognitive modelling with those of didactic instruction on children’s long-division self-efficacy and achievement. Children lacking division skills received instruction and practice. In the cognitive modelling condition, students observed an adult model explain and demonstrate division operations whilst applying them to sample problems. In the didactic instruction condition, students reviewed instructional material that explained and demonstrated the operations, but they were not exposed to models. Cognitive modelling enhanced children’s division achievement better than did didactic instruction.
Teacher Modelling
Teachers oft incorporate modelled demonstrations into lessons designed to teach students diverse skills such as solving mathematical problems, identifying main ideas in text, writing topic sentences, using power tools, and executing defensive basketball manoeuvres. Modelled demonstrations can be used to teach elementary school children how to head their papers properly. In her third-grade class, Kathy Stone might draw on the board a sketch of the paper students are using. She then can review the heading procedure step by step, explaining and demonstrating how to complete it.
In his ninth-grade American history class, Jim Marshall models how to study for a test. Working through several chapters, he explains and demonstrates how to locate and summarise the major terms and points for each section.
In a middle school life skills class, students can learn how to insert a sleeve into a garment through modelled demonstrations. The teacher might begin by describing the process and then use visual aids to portray the procedure. The teacher could conclude the presentation by demonstrating the process at a sewing machine.
Several students in Gina Brown’s undergraduate class have been coming to her office after class with questions about how to present their findings from their field projects. During the next class, she uses a research project she completed to demonstrate how one might present findings to a group. She uses handouts, charts, and PowerPoint® to illustrate ways to present data.
A drama teacher can model various performance skills whilst working with students as they practice a play. The teacher can demonstrate desired voice inflections, mood, volume, and body movements for each character in the play. Whilst presenting a word decoding lesson using phonics, a first-grade teacher can demonstrate sounding out each letter in a list of words.
Self-Instruction
Self-instruction hath been used to teach students to regulate their activities during learning (Meichenbaum, 1977). In an early study, Meichenbaum and Goodman (1971) incorporated cognitive modelling into self-instructional training with impulsive second graders in a special-education class. The procedure included:
- Cognitive modelling: Adult tells child what to do whilst adult performs the task.
- Overt guidance: Child performs under direction of adult.
- Overt self-guidance: Child performs whilst self-instructing aloud.
- Faded overt self-guidance: Child whispers instructions whilst performing task.
- Covert self-instruction: Child performs whilst guided by inner silent speech.
Self-instruction oft is used to slow down children’s rate of performing. An adult model used the following statements during a line-drawing task:
Meichenbaum & Goodman (1971, p. 117):
Okay, what is it I have to do? You want me to copy the picture with the different lines. I have to go slow and be careful. Okay, draw the line down, down, good; then to the right, that’s it; now down some more and to the left. Good, I’m doing fine so far. Remember go slow. Now back up again. Nay, I was supposed to go down. That’s okay, just erase the line carefully. . . . Good. Even if I make an error I can go on slowly and carefully. Okay, I have to go down now. Finished. I did it.
Note that the model maketh a mistake and showeth how to deal with it. This is an important form of learning for students with attention-deficit disorders, hyperactivity, and behavioural problems because they may become frustrated and quit easily following errors. Meichenbaum and Goodman (1971) found that cognitive modelling slowed down response times, but that the self-instructions decreased errors.
Self-instruction hath been used with a variety of tasks and types of students (Fish & Pervan, 1985). It is especially useful for students with learning disabilities (Wood, Rosenberg, & Carran, 1993) and for teaching students to work strategically. In teaching reading comprehension, the preceding instructions might be modified as follows: “What is it I have to do? I have to find the topic sentence of the paragraph. The topic sentence is what the paragraph is about. I start by looking for a sentence that sums up the details or tells what the paragraph is about” (McNeil, 1987, p. 96). Statements for coping with difficulties (“I haven’t found it yet, but that’s all right”) can be built into the modelled demonstration.
Motor Skill Learning
Social cognitive theory doth posit that motor skill learning involveth the construction of a mental model. This model provideth a conceptual representation of the skill, serving as the standard for the correction of responses subsequent to the receipt of feedback (Bandura, 1986; McCullagh, 1993; Weiss, Ebbeck, & Wiese-Bjornstal, 1993). This conceptual representation is formed by the transformation of observed sequences of behaviours into visual and symbolic codes, which are then cognitively rehearsed. Individuals oft possess a mental model of a skill before attempting its performance. For instance, by observing tennis players, individuals construct a mental model of activities such as the serve, volley, and back-hand. These models, though rudimentary, require feedback and correction for their perfection, yet allow learners to perform approximations of the skills from the outset of training. This phenomenon was witnessed in the opening scenario, wherein Donnetta was required to construct a mental model of a backhand. In the case of novel or complex behaviours, learners may lack a prior mental model and must needs observe modelled demonstrations before attempting the behaviours.
The social cognitive approach to motor skill learning doth diverge from traditional explanations. Adams's (1971) closed-loop theory doth postulate that individuals develop perceptual (internal) traces of motor skill movements through practice and feedback. These traces serve as the reference for correct movements. As one performeth a behaviour, one receiveth internal (sensory) and external (knowledge of results) feedback, and compareth this feedback to the trace. Any discrepancy serveth to correct the trace. Learning is enhanced when feedback is accurate, and eventually the behaviour can be performed without feedback. Adams didst distinguish two memory mechanisms: one for producing the response, and another for evaluating its correctness.
A differing view is grounded in schema theory (Schmidt, 1975). Schmidt didst postulate that individuals store in memory a wealth of information regarding motor skill movements, including the initial conditions, the characteristics of the generalized motor sequence, the results of the movement, knowledge of results, and sensory feedback. Learners store this information within two general schemas, or organized memory networks comprising related information. The recall schema concerneth response production, whilst the recognition schema is employed to evaluate responses.
Social cognitive theory maintaineth that by observing others, individuals form a cognitive representation that initiateth subsequent responses and serveth as a standard for the evaluation of the correctness of said responses (Bandura, 1986). Motor learning theories differ from social cognitive theory in that they place greater emphasis on error correction after acting, and postulate two memory mechanisms for the storage of information and the evaluation of accuracy (McCullagh, 1993). Social cognitive theory also doth highlight the role of personal cognitions (goals and expectations) in the development of motor skills.
A challenge in motor skill learning lieth in the learner's inability to observe aspects of their performances that fall outside their field of vision. Individuals swinging a golf club, hitting a tennis serve, kicking a football, throwing a baseball, or hurling a discus, cannot observe many aspects of these sequences. This inability to see one's actions requireth reliance on kinesthetic feedback, which must be compared to one's conceptual representation. The absence of visual feedback doth render learning difficult.
Carroll and Bandura (1982) didst expose learners to models performing a motor skill, and then requested that they reproduce the motor pattern. The experimenters didst provide some learners with concurrent visual feedback of their performances, using a video camera to allow them to observe their real-time performances on a monitor. Other learners didst not receive visual feedback. When visual feedback was provided before learners had formed a mental model of the motor behaviour, it exerted no effect on performance. However, once learners possessed an adequate model, visual feedback enhanced their accurate reproduction of the modelled behaviours, eliminating discrepancies between their conceptual models and their actions.
Motor Skill Learning
Observational learning proveth useful in the acquisition of motor skills. To instruct students in the art of dribbling a basketball, physical education teachers commence with skill exercises, such as standing stationary and bouncing the ball, and moving whilst bouncing the ball with each step. After introducing each skill leading to the final sequence, teachers can demonstrate slowly and precisely what the students are to model. The students should then practice that skill. Should students encounter difficulty at a particular step, teachers can repeat the modelled demonstration before the students continue practicing.
For high school students to successfully learn a dance for performance in the spring musical, the teacher must needs demonstrate and slowly progress toward setting the dance to music. The teacher may divide the dance, working on each step separately, gradually combining steps, and eventually uniting all the various steps with the music.
Researchers have also examined the efficacy of employing models to teach motor skills. Weiss (1983) didst compare the effects of a silent model (visual demonstration) with those of a verbal model (visual demonstration plus verbal explanation) on the learning of a six-part motor skill obstacle course. Older children (ages 7 through 9 years) learned equally well with either model; younger children (ages 4 through 6 years) learned better with the verbal model. Perchance the addition of verbalizations created a cognitive model that aided in maintaining children's attention and assisted with the coding of information in memory. Weiss and Klint (1987) discovered that children in visual-model and no-model conditions who verbally rehearsed the sequence of actions learned the motor skills better than children who did not verbally rehearse. Collectively, these results suggest that some form of verbalization may be critically important in the acquisition of motor skills.